mirror of
				git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git
				synced 2025-10-31 08:44:41 +00:00 
			
		
		
		
	 3270476a6c
			
		
	
	
		3270476a6c
		
	
	
	
	
		
			
			WQ_HIGHPRI was implemented by queueing highpri work items at the head
of the global worklist.  Other than queueing at the head, they weren't
handled differently; unfortunately, this could lead to execution
latency of a few seconds on heavily loaded systems.
Now that workqueue code has been updated to deal with multiple
worker_pools per global_cwq, this patch reimplements WQ_HIGHPRI using
a separate worker_pool.  NR_WORKER_POOLS is bumped to two and
gcwq->pools[0] is used for normal pri work items and ->pools[1] for
highpri.  Highpri workers get -20 nice level and has 'H' suffix in
their names.  Note that this change increases the number of kworkers
per cpu.
POOL_HIGHPRI_PENDING, pool_determine_ins_pos() and highpri chain
wakeup code in process_one_work() are no longer used and removed.
This allows proper prioritization of highpri work items and removes
high execution latency of highpri work items.
v2: nr_running indexing bug in get_pool_nr_running() fixed.
v3: Refreshed for the get_pool_nr_running() update in the previous
    patch.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reported-by: Josh Hunt <joshhunt00@gmail.com>
LKML-Reference: <CAKA=qzaHqwZ8eqpLNFjxnO2fX-tgAOjmpvxgBFjv6dJeQaOW1w@mail.gmail.com>
Cc: Tony Luck <tony.luck@intel.com>
Cc: Fengguang Wu <fengguang.wu@intel.com>
		
	
			
		
			
				
	
	
		
			394 lines
		
	
	
	
		
			15 KiB
		
	
	
	
		
			Text
		
	
	
	
	
	
			
		
		
	
	
			394 lines
		
	
	
	
		
			15 KiB
		
	
	
	
		
			Text
		
	
	
	
	
	
| 
 | |
| Concurrency Managed Workqueue (cmwq)
 | |
| 
 | |
| September, 2010		Tejun Heo <tj@kernel.org>
 | |
| 			Florian Mickler <florian@mickler.org>
 | |
| 
 | |
| CONTENTS
 | |
| 
 | |
| 1. Introduction
 | |
| 2. Why cmwq?
 | |
| 3. The Design
 | |
| 4. Application Programming Interface (API)
 | |
| 5. Example Execution Scenarios
 | |
| 6. Guidelines
 | |
| 7. Debugging
 | |
| 
 | |
| 
 | |
| 1. Introduction
 | |
| 
 | |
| There are many cases where an asynchronous process execution context
 | |
| is needed and the workqueue (wq) API is the most commonly used
 | |
| mechanism for such cases.
 | |
| 
 | |
| When such an asynchronous execution context is needed, a work item
 | |
| describing which function to execute is put on a queue.  An
 | |
| independent thread serves as the asynchronous execution context.  The
 | |
| queue is called workqueue and the thread is called worker.
 | |
| 
 | |
| While there are work items on the workqueue the worker executes the
 | |
| functions associated with the work items one after the other.  When
 | |
| there is no work item left on the workqueue the worker becomes idle.
 | |
| When a new work item gets queued, the worker begins executing again.
 | |
| 
 | |
| 
 | |
| 2. Why cmwq?
 | |
| 
 | |
| In the original wq implementation, a multi threaded (MT) wq had one
 | |
| worker thread per CPU and a single threaded (ST) wq had one worker
 | |
| thread system-wide.  A single MT wq needed to keep around the same
 | |
| number of workers as the number of CPUs.  The kernel grew a lot of MT
 | |
| wq users over the years and with the number of CPU cores continuously
 | |
| rising, some systems saturated the default 32k PID space just booting
 | |
| up.
 | |
| 
 | |
| Although MT wq wasted a lot of resource, the level of concurrency
 | |
| provided was unsatisfactory.  The limitation was common to both ST and
 | |
| MT wq albeit less severe on MT.  Each wq maintained its own separate
 | |
| worker pool.  A MT wq could provide only one execution context per CPU
 | |
| while a ST wq one for the whole system.  Work items had to compete for
 | |
| those very limited execution contexts leading to various problems
 | |
| including proneness to deadlocks around the single execution context.
 | |
| 
 | |
| The tension between the provided level of concurrency and resource
 | |
| usage also forced its users to make unnecessary tradeoffs like libata
 | |
| choosing to use ST wq for polling PIOs and accepting an unnecessary
 | |
| limitation that no two polling PIOs can progress at the same time.  As
 | |
| MT wq don't provide much better concurrency, users which require
 | |
| higher level of concurrency, like async or fscache, had to implement
 | |
| their own thread pool.
 | |
| 
 | |
| Concurrency Managed Workqueue (cmwq) is a reimplementation of wq with
 | |
| focus on the following goals.
 | |
| 
 | |
| * Maintain compatibility with the original workqueue API.
 | |
| 
 | |
| * Use per-CPU unified worker pools shared by all wq to provide
 | |
|   flexible level of concurrency on demand without wasting a lot of
 | |
|   resource.
 | |
| 
 | |
| * Automatically regulate worker pool and level of concurrency so that
 | |
|   the API users don't need to worry about such details.
 | |
| 
 | |
| 
 | |
| 3. The Design
 | |
| 
 | |
| In order to ease the asynchronous execution of functions a new
 | |
| abstraction, the work item, is introduced.
 | |
| 
 | |
| A work item is a simple struct that holds a pointer to the function
 | |
| that is to be executed asynchronously.  Whenever a driver or subsystem
 | |
| wants a function to be executed asynchronously it has to set up a work
 | |
| item pointing to that function and queue that work item on a
 | |
| workqueue.
 | |
| 
 | |
| Special purpose threads, called worker threads, execute the functions
 | |
| off of the queue, one after the other.  If no work is queued, the
 | |
| worker threads become idle.  These worker threads are managed in so
 | |
| called thread-pools.
 | |
| 
 | |
| The cmwq design differentiates between the user-facing workqueues that
 | |
| subsystems and drivers queue work items on and the backend mechanism
 | |
| which manages thread-pools and processes the queued work items.
 | |
| 
 | |
| The backend is called gcwq.  There is one gcwq for each possible CPU
 | |
| and one gcwq to serve work items queued on unbound workqueues.  Each
 | |
| gcwq has two thread-pools - one for normal work items and the other
 | |
| for high priority ones.
 | |
| 
 | |
| Subsystems and drivers can create and queue work items through special
 | |
| workqueue API functions as they see fit. They can influence some
 | |
| aspects of the way the work items are executed by setting flags on the
 | |
| workqueue they are putting the work item on. These flags include
 | |
| things like CPU locality, reentrancy, concurrency limits, priority and
 | |
| more.  To get a detailed overview refer to the API description of
 | |
| alloc_workqueue() below.
 | |
| 
 | |
| When a work item is queued to a workqueue, the target gcwq and
 | |
| thread-pool is determined according to the queue parameters and
 | |
| workqueue attributes and appended on the shared worklist of the
 | |
| thread-pool.  For example, unless specifically overridden, a work item
 | |
| of a bound workqueue will be queued on the worklist of either normal
 | |
| or highpri thread-pool of the gcwq that is associated to the CPU the
 | |
| issuer is running on.
 | |
| 
 | |
| For any worker pool implementation, managing the concurrency level
 | |
| (how many execution contexts are active) is an important issue.  cmwq
 | |
| tries to keep the concurrency at a minimal but sufficient level.
 | |
| Minimal to save resources and sufficient in that the system is used at
 | |
| its full capacity.
 | |
| 
 | |
| Each thread-pool bound to an actual CPU implements concurrency
 | |
| management by hooking into the scheduler.  The thread-pool is notified
 | |
| whenever an active worker wakes up or sleeps and keeps track of the
 | |
| number of the currently runnable workers.  Generally, work items are
 | |
| not expected to hog a CPU and consume many cycles.  That means
 | |
| maintaining just enough concurrency to prevent work processing from
 | |
| stalling should be optimal.  As long as there are one or more runnable
 | |
| workers on the CPU, the thread-pool doesn't start execution of a new
 | |
| work, but, when the last running worker goes to sleep, it immediately
 | |
| schedules a new worker so that the CPU doesn't sit idle while there
 | |
| are pending work items.  This allows using a minimal number of workers
 | |
| without losing execution bandwidth.
 | |
| 
 | |
| Keeping idle workers around doesn't cost other than the memory space
 | |
| for kthreads, so cmwq holds onto idle ones for a while before killing
 | |
| them.
 | |
| 
 | |
| For an unbound wq, the above concurrency management doesn't apply and
 | |
| the thread-pools for the pseudo unbound CPU try to start executing all
 | |
| work items as soon as possible.  The responsibility of regulating
 | |
| concurrency level is on the users.  There is also a flag to mark a
 | |
| bound wq to ignore the concurrency management.  Please refer to the
 | |
| API section for details.
 | |
| 
 | |
| Forward progress guarantee relies on that workers can be created when
 | |
| more execution contexts are necessary, which in turn is guaranteed
 | |
| through the use of rescue workers.  All work items which might be used
 | |
| on code paths that handle memory reclaim are required to be queued on
 | |
| wq's that have a rescue-worker reserved for execution under memory
 | |
| pressure.  Else it is possible that the thread-pool deadlocks waiting
 | |
| for execution contexts to free up.
 | |
| 
 | |
| 
 | |
| 4. Application Programming Interface (API)
 | |
| 
 | |
| alloc_workqueue() allocates a wq.  The original create_*workqueue()
 | |
| functions are deprecated and scheduled for removal.  alloc_workqueue()
 | |
| takes three arguments - @name, @flags and @max_active.  @name is the
 | |
| name of the wq and also used as the name of the rescuer thread if
 | |
| there is one.
 | |
| 
 | |
| A wq no longer manages execution resources but serves as a domain for
 | |
| forward progress guarantee, flush and work item attributes.  @flags
 | |
| and @max_active control how work items are assigned execution
 | |
| resources, scheduled and executed.
 | |
| 
 | |
| @flags:
 | |
| 
 | |
|   WQ_NON_REENTRANT
 | |
| 
 | |
| 	By default, a wq guarantees non-reentrance only on the same
 | |
| 	CPU.  A work item may not be executed concurrently on the same
 | |
| 	CPU by multiple workers but is allowed to be executed
 | |
| 	concurrently on multiple CPUs.  This flag makes sure
 | |
| 	non-reentrance is enforced across all CPUs.  Work items queued
 | |
| 	to a non-reentrant wq are guaranteed to be executed by at most
 | |
| 	one worker system-wide at any given time.
 | |
| 
 | |
|   WQ_UNBOUND
 | |
| 
 | |
| 	Work items queued to an unbound wq are served by a special
 | |
| 	gcwq which hosts workers which are not bound to any specific
 | |
| 	CPU.  This makes the wq behave as a simple execution context
 | |
| 	provider without concurrency management.  The unbound gcwq
 | |
| 	tries to start execution of work items as soon as possible.
 | |
| 	Unbound wq sacrifices locality but is useful for the following
 | |
| 	cases.
 | |
| 
 | |
| 	* Wide fluctuation in the concurrency level requirement is
 | |
| 	  expected and using bound wq may end up creating large number
 | |
| 	  of mostly unused workers across different CPUs as the issuer
 | |
| 	  hops through different CPUs.
 | |
| 
 | |
| 	* Long running CPU intensive workloads which can be better
 | |
| 	  managed by the system scheduler.
 | |
| 
 | |
|   WQ_FREEZABLE
 | |
| 
 | |
| 	A freezable wq participates in the freeze phase of the system
 | |
| 	suspend operations.  Work items on the wq are drained and no
 | |
| 	new work item starts execution until thawed.
 | |
| 
 | |
|   WQ_MEM_RECLAIM
 | |
| 
 | |
| 	All wq which might be used in the memory reclaim paths _MUST_
 | |
| 	have this flag set.  The wq is guaranteed to have at least one
 | |
| 	execution context regardless of memory pressure.
 | |
| 
 | |
|   WQ_HIGHPRI
 | |
| 
 | |
| 	Work items of a highpri wq are queued to the highpri
 | |
| 	thread-pool of the target gcwq.  Highpri thread-pools are
 | |
| 	served by worker threads with elevated nice level.
 | |
| 
 | |
| 	Note that normal and highpri thread-pools don't interact with
 | |
| 	each other.  Each maintain its separate pool of workers and
 | |
| 	implements concurrency management among its workers.
 | |
| 
 | |
|   WQ_CPU_INTENSIVE
 | |
| 
 | |
| 	Work items of a CPU intensive wq do not contribute to the
 | |
| 	concurrency level.  In other words, runnable CPU intensive
 | |
| 	work items will not prevent other work items in the same
 | |
| 	thread-pool from starting execution.  This is useful for bound
 | |
| 	work items which are expected to hog CPU cycles so that their
 | |
| 	execution is regulated by the system scheduler.
 | |
| 
 | |
| 	Although CPU intensive work items don't contribute to the
 | |
| 	concurrency level, start of their executions is still
 | |
| 	regulated by the concurrency management and runnable
 | |
| 	non-CPU-intensive work items can delay execution of CPU
 | |
| 	intensive work items.
 | |
| 
 | |
| 	This flag is meaningless for unbound wq.
 | |
| 
 | |
| @max_active:
 | |
| 
 | |
| @max_active determines the maximum number of execution contexts per
 | |
| CPU which can be assigned to the work items of a wq.  For example,
 | |
| with @max_active of 16, at most 16 work items of the wq can be
 | |
| executing at the same time per CPU.
 | |
| 
 | |
| Currently, for a bound wq, the maximum limit for @max_active is 512
 | |
| and the default value used when 0 is specified is 256.  For an unbound
 | |
| wq, the limit is higher of 512 and 4 * num_possible_cpus().  These
 | |
| values are chosen sufficiently high such that they are not the
 | |
| limiting factor while providing protection in runaway cases.
 | |
| 
 | |
| The number of active work items of a wq is usually regulated by the
 | |
| users of the wq, more specifically, by how many work items the users
 | |
| may queue at the same time.  Unless there is a specific need for
 | |
| throttling the number of active work items, specifying '0' is
 | |
| recommended.
 | |
| 
 | |
| Some users depend on the strict execution ordering of ST wq.  The
 | |
| combination of @max_active of 1 and WQ_UNBOUND is used to achieve this
 | |
| behavior.  Work items on such wq are always queued to the unbound gcwq
 | |
| and only one work item can be active at any given time thus achieving
 | |
| the same ordering property as ST wq.
 | |
| 
 | |
| 
 | |
| 5. Example Execution Scenarios
 | |
| 
 | |
| The following example execution scenarios try to illustrate how cmwq
 | |
| behave under different configurations.
 | |
| 
 | |
|  Work items w0, w1, w2 are queued to a bound wq q0 on the same CPU.
 | |
|  w0 burns CPU for 5ms then sleeps for 10ms then burns CPU for 5ms
 | |
|  again before finishing.  w1 and w2 burn CPU for 5ms then sleep for
 | |
|  10ms.
 | |
| 
 | |
| Ignoring all other tasks, works and processing overhead, and assuming
 | |
| simple FIFO scheduling, the following is one highly simplified version
 | |
| of possible sequences of events with the original wq.
 | |
| 
 | |
|  TIME IN MSECS	EVENT
 | |
|  0		w0 starts and burns CPU
 | |
|  5		w0 sleeps
 | |
|  15		w0 wakes up and burns CPU
 | |
|  20		w0 finishes
 | |
|  20		w1 starts and burns CPU
 | |
|  25		w1 sleeps
 | |
|  35		w1 wakes up and finishes
 | |
|  35		w2 starts and burns CPU
 | |
|  40		w2 sleeps
 | |
|  50		w2 wakes up and finishes
 | |
| 
 | |
| And with cmwq with @max_active >= 3,
 | |
| 
 | |
|  TIME IN MSECS	EVENT
 | |
|  0		w0 starts and burns CPU
 | |
|  5		w0 sleeps
 | |
|  5		w1 starts and burns CPU
 | |
|  10		w1 sleeps
 | |
|  10		w2 starts and burns CPU
 | |
|  15		w2 sleeps
 | |
|  15		w0 wakes up and burns CPU
 | |
|  20		w0 finishes
 | |
|  20		w1 wakes up and finishes
 | |
|  25		w2 wakes up and finishes
 | |
| 
 | |
| If @max_active == 2,
 | |
| 
 | |
|  TIME IN MSECS	EVENT
 | |
|  0		w0 starts and burns CPU
 | |
|  5		w0 sleeps
 | |
|  5		w1 starts and burns CPU
 | |
|  10		w1 sleeps
 | |
|  15		w0 wakes up and burns CPU
 | |
|  20		w0 finishes
 | |
|  20		w1 wakes up and finishes
 | |
|  20		w2 starts and burns CPU
 | |
|  25		w2 sleeps
 | |
|  35		w2 wakes up and finishes
 | |
| 
 | |
| Now, let's assume w1 and w2 are queued to a different wq q1 which has
 | |
| WQ_CPU_INTENSIVE set,
 | |
| 
 | |
|  TIME IN MSECS	EVENT
 | |
|  0		w0 starts and burns CPU
 | |
|  5		w0 sleeps
 | |
|  5		w1 and w2 start and burn CPU
 | |
|  10		w1 sleeps
 | |
|  15		w2 sleeps
 | |
|  15		w0 wakes up and burns CPU
 | |
|  20		w0 finishes
 | |
|  20		w1 wakes up and finishes
 | |
|  25		w2 wakes up and finishes
 | |
| 
 | |
| 
 | |
| 6. Guidelines
 | |
| 
 | |
| * Do not forget to use WQ_MEM_RECLAIM if a wq may process work items
 | |
|   which are used during memory reclaim.  Each wq with WQ_MEM_RECLAIM
 | |
|   set has an execution context reserved for it.  If there is
 | |
|   dependency among multiple work items used during memory reclaim,
 | |
|   they should be queued to separate wq each with WQ_MEM_RECLAIM.
 | |
| 
 | |
| * Unless strict ordering is required, there is no need to use ST wq.
 | |
| 
 | |
| * Unless there is a specific need, using 0 for @max_active is
 | |
|   recommended.  In most use cases, concurrency level usually stays
 | |
|   well under the default limit.
 | |
| 
 | |
| * A wq serves as a domain for forward progress guarantee
 | |
|   (WQ_MEM_RECLAIM, flush and work item attributes.  Work items which
 | |
|   are not involved in memory reclaim and don't need to be flushed as a
 | |
|   part of a group of work items, and don't require any special
 | |
|   attribute, can use one of the system wq.  There is no difference in
 | |
|   execution characteristics between using a dedicated wq and a system
 | |
|   wq.
 | |
| 
 | |
| * Unless work items are expected to consume a huge amount of CPU
 | |
|   cycles, using a bound wq is usually beneficial due to the increased
 | |
|   level of locality in wq operations and work item execution.
 | |
| 
 | |
| 
 | |
| 7. Debugging
 | |
| 
 | |
| Because the work functions are executed by generic worker threads
 | |
| there are a few tricks needed to shed some light on misbehaving
 | |
| workqueue users.
 | |
| 
 | |
| Worker threads show up in the process list as:
 | |
| 
 | |
| root      5671  0.0  0.0      0     0 ?        S    12:07   0:00 [kworker/0:1]
 | |
| root      5672  0.0  0.0      0     0 ?        S    12:07   0:00 [kworker/1:2]
 | |
| root      5673  0.0  0.0      0     0 ?        S    12:12   0:00 [kworker/0:0]
 | |
| root      5674  0.0  0.0      0     0 ?        S    12:13   0:00 [kworker/1:0]
 | |
| 
 | |
| If kworkers are going crazy (using too much cpu), there are two types
 | |
| of possible problems:
 | |
| 
 | |
| 	1. Something beeing scheduled in rapid succession
 | |
| 	2. A single work item that consumes lots of cpu cycles
 | |
| 
 | |
| The first one can be tracked using tracing:
 | |
| 
 | |
| 	$ echo workqueue:workqueue_queue_work > /sys/kernel/debug/tracing/set_event
 | |
| 	$ cat /sys/kernel/debug/tracing/trace_pipe > out.txt
 | |
| 	(wait a few secs)
 | |
| 	^C
 | |
| 
 | |
| If something is busy looping on work queueing, it would be dominating
 | |
| the output and the offender can be determined with the work item
 | |
| function.
 | |
| 
 | |
| For the second type of problems it should be possible to just check
 | |
| the stack trace of the offending worker thread.
 | |
| 
 | |
| 	$ cat /proc/THE_OFFENDING_KWORKER/stack
 | |
| 
 | |
| The work item's function should be trivially visible in the stack
 | |
| trace.
 |